Coordinated Operation and Control in Energy Internet Using Machine Learning and Modeling Optimization

As machine learning (artificial neural network) and system modeling optimization each has individual advantages and disadvantages, some functions are complemented and can make up different shortcomings, so integrated using them may achieve performance revenues in system optimization and management. In the Energy Internet (EI) managing and control domain, we sequentially using ELM (extreme learning machine) for machine learning and PSO (particle swarm optimization) for system modeling optimization which should obey inherent operation constraints, to solve the system optimizing problem. This algorithm is tested in EI with one day before operation, and its performance is compared with that of pure PSO algorithm. The results show great running time reduction and negligible running performance degradation, and can be effectively used in cloud-edge collaboration of EI system.